Page 7

Dear Mr. Double,

I have received your e-mail on the subject of slenderness of struts in
BS:5950 Pt1-2000. I am sorry for this belated response.

I am a Chartered Structural Engineer from India, and have worked in UK from
1964-1977. I have used BS 449 while I was in UK. The present Indian code
IS:800-1984 for steelwork is similar to BS 449. From BS 449, we all knew
very well how useful the slenderness limits are while designing struts. The
proposed draft revision to IS:800 which I believe is similar to BS:5950
Pt1-2000, gives stress reduction factors for buckling for Kl/r values upto
250. Similarly, the American AISC manual (9th Edn.) for allowable Stress
Design gives permissible values of axial stress for Kl/r values upto 200
only.

I have been following the Verulam columns on the above subject, and I fully
agree with you that the aspect of slenderness limit should be re-instated in
BS:5950 Pt1-2000, as otherwise there could be mis-use or abuse of the code.
In the present day usage of computers, how many designers fully understand
the implications P-delta analysis while designing struts? How many
designers consider second order effects while designing a simple strut in
bracings or roof truss? These aspects have been taken up very well in your
Verulam letter of 4th May 2004, and also in the letter by Gary Wyatt, Dr.
Dennis Lam and Prof. Siu Lai Chan, published in the Verulam column of
21.9.2004. These are some of the ponderable questions which the code
committee should think about while preparing or revising the code.

Yours truly,

N. P

C S

_______________________________________________________________________________________

In my experience of lightly loaded building structures, roof trusses etc. the slenderness limits often introduced an additional robustness into the structure.

In view of the concerns raised I suggest that the BSI should draw the Designer's attention to the fact that exceeding (the old) slenderness ratios may lead to a less than robust member/structure.

R H

_______________________________________________________________________________________

Martin ,

When designing struts I always choose size which satisfies l/ry 180 as a minimum, this usually is sufficient to carry the load.

Regards

A W

_______________________________________________________________________________________

Hi Martin,

Sorry for not responding sooner.

a) I do not support BS5950 as it stands

b) I think it is essential to reinstate L/r 's

The bulk of my design is to SABS 0162 (South African Bureau of Standards) which is loosely based on the Canadian Code. I am on the technical committee of the Southern African Institute of Steel Construction who have recently revisited the SABS 0162 content & we have retained the slenderness limits .

My question to you is why are amendments being made to BS 5950 when Euro codes should soon be in use ?

regards D N

_______________________________________________________________________________________

Good morning Martin,

I apologise for the delay in replying but I have been catching up with the story so far.
My vote is a very definite (b).
I am also of the view that BS 449 should be retained for those small simple designs
that require the odd column or beam.

Keep up the good work.
Kind regards
C Y

_______________________________________________________________________________________

Neil,

Many thanks for the publication, P344, which I have had a good look through. The guidance on structural design, building layout, service integration etc is all good stuff, I would therefore recommend the publication to graduates and to other non-steelwork-specialists.

The section (7.0) on frame stability indicates how frame stability checks should be carried out in accordance with BS5950. This in itself is all very well except that there is a problem with BS5950 in this respect:-

* BS5950-2000 fails to include "member P.d effects" in its recommendations for frame sway sensitivity.

* With BS5950-1990, there was some control over the slenderness of strut members. In this case "member P.d effects" were also controlled and might therefore have been considered trivial, and having no effect on frame stability.

* Now that BS5950-2000 has dropped strut slenderness limits it is possible for frame stability to be affected grossly by "member P.d effects"....etc

Regards

Martin Double
CADOSS

_______________________________________________________________________________________

Dear Martin,

I was wondering if you have received any responses from some structural engineers in Hong Kong. I have actually forwarded your missive to a handful of engineers in Hong Kong who might have experience on the subject matter. If you are still interested in the subject, I would not mind to 'reach out' again. I must apologize for not responding to you direct because I do not speak the language, so to speak!

Regards,

F S H NG

_______________________________________________________________________________________

Dear Martin,

Most people using a design code expect some guidance rather than just the
basic principles.

I would support reinstatement of slenderness limits or similar (i.e. Option
b). If BSI feels uneasy about giving guidance on an issue that is too
complicated to be represented by just a few parameters, the limitations on
the use of such slenderness limits may be stated together with some kind of
warning. The slenderness limits may be set conservatively. The code may
specify that any struts with slenderness values above these limits are
required to go through whatever rigorous analysis.

Regards,

FTKA

_______________________________________________________________________________________

Dear engineers,

I refer to the question raised by Martin below for opinions of structural engineers.

The Department of Civil & Structural Engineering of the Hong Kong Polytechnic University, jointly with Ove Arup and Partners, is involved in the code drafting of the coming steel code in Hong Kong and we considered the issue of limitation of slenderness . In fact, the article mentioned by Martin below and appeared in Verulam of the IStructE journal is written by myself, together with friends in U.K. but Mr. Gary's Wyatt's and my affiliations were incorrectly printed as Leeds University.

With the use of computational method, the limitation is no longer considered as necessary, PROVIDED that the effect due to load (self-weight, wind or snow loads) along the member and its subsequent P-d effect is considered. Otherwise, if this P-d effect is ignored (i.e. when using a linear analysis or P-D-only analysis), we should then include the restriction to limit the ignored effect of P-d moment induced by these loads through maximum slenderness ratio. In summary, we believe that the code should have limitation when one uses a linear or P-D-only analysis, which ignores the effect, but not for a non-linear P-D-d analysis which has already considered the effect of high slenderness.

Note : P-D (big D) effect refers to change of joint geometry and P-d (small d) effect refers to change of member bow.

S.L. C

_______________________________________________________________________________________

Dear Martin,

On the issue of slender struts, I would like to add my support for the
reinstatement of slenderness limits applicable to strut design, item b) in
your categories.

I am fortunate enough that I can start to comprehend second order effects
and the problems generated, however as a checking engineer, I come across
instances where struts have exceeded slenderness limits with no
consideration given to second order effects. It is my view that the
slenderness limits should be re-instated as a minimum requirement with a
proviso that they can be increased if the design takes account of second
order effects.

regards Simon.

_______________________________________________________________________________________

Dear Martin

As yet the issue of strut slenderness has not reared its head in our office. Largely because we do not deal with significant structural steel structures on a regular basis. However, I am always sceptical about the general approach where if the computer says its right it is. I have had a number of cases in the past where engineers have modelled a structure without fully understanding its behaviour and in almost, most notably where using finite elements, all cases these have lead to some potentially serious shortfalls in the final design.

Like a number of others who have responded I also believe that codes should not be prescriptive but a guide to good practice for both the experienced and inexperienced. Consequently, I consider that appropriate guidelines on the maximum slenderness should be re-introduced, and will continue to use the earlier well established rules until such time as either something better or a more detailed explanation as to why they can be omitted is published.

Regards

Phil

_______________________________________________________________________________________

Hi Martin,

Sorry for this very late reply (have been in Dubai and could not access my ISP).

My opinion is b) Support for reinstatement of slenderness limits or similar

Kind regards,

Arshad

_______________________________________________________________________________________

my response is ( b )

leave well enough alone

P C

_______________________________________________________________________________________